upgini 1.2.113a3__py3-none-any.whl → 1.2.113a5__py3-none-any.whl

upgini/__about__.py

@@ -1 +1 @@
-__version__ = "1.2.113a3"
+__version__ = "1.2.113a5"

upgini/features_enricher.py

@@ -112,7 +112,7 @@ except Exception:
         CustomFallbackProgressBar as ProgressBar,
     )
 
-from upgini.utils.psi import calculate_features_psi
+from upgini.utils.psi import calculate_features_psi, calculate_sparsity_psi
 from upgini.utils.sample_utils import SampleColumns, SampleConfig, _num_samples, sample
 from upgini.utils.sort import sort_columns
 from upgini.utils.target_utils import calculate_psi, define_task
@@ -1059,24 +1059,9 @@ class FeaturesEnricher(TransformerMixin):
                     groups,
                     _cv,
                     columns_renaming,
-                    eval_set_dates,
+                    _,
                 ) = prepared_data
 
-                # rename cat_features
-                if client_cat_features:
-                    for new_c, old_c in columns_renaming.items():
-                        if old_c in client_cat_features:
-                            client_cat_features.remove(old_c)
-                            client_cat_features.append(new_c)
-                    for cat_feature in client_cat_features:
-                        if cat_feature not in fitting_X.columns:
-                            self.logger.error(
-                                f"Client cat_feature `{cat_feature}` not found in"
-                                f" x columns: {fitting_X.columns.to_list()}"
-                            )
-                else:
-                    client_cat_features = []
-
                 # rename baseline_score_column
                 reversed_renaming = {v: k for k, v in columns_renaming.items()}
                 baseline_score_column = self.baseline_score_column
@@ -1305,7 +1290,7 @@ class FeaturesEnricher(TransformerMixin):
                             metrics.append(eval_metrics)
 
                     if updating_shaps is not None:
-                        decoded_X = self._decode_id_columns(fitting_X, columns_renaming)
+                        decoded_X = self._decode_id_columns(fitting_X)
                         self._update_shap_values(trace_id, decoded_X, updating_shaps, silent=not internal_call)
 
                     metrics_df = pd.DataFrame(metrics)
@@ -1376,12 +1361,23 @@ class FeaturesEnricher(TransformerMixin):
         if isinstance(X, np.ndarray):
             search_keys = {str(k): v for k, v in search_keys.items()}
 
-        has_date = self._get_date_column(search_keys) is not None
-        if not has_date or not validated_eval_set:
-            self.logger.info("No date column or eval set for OOT psi calculation")
+        date_column = self._get_date_column(search_keys)
+        has_date = date_column is not None
+        if not has_date:
+            self.logger.info("No date column for OOT PSI calculation")
+            return
+        if not validated_eval_set:
+            self.logger.info("No eval set for OOT PSI calculation")
+            return
+        if validated_X[date_column].nunique() <= 1:
+            self.logger.warning("Constant date for OOT PSI calculation")
+            return
+        if self.cv is not None and self.cv.is_time_series():
+            self.logger.warning("Time series CV is not supported for OOT PSI calculation")
             return
 
         cat_features_from_backend = self.__get_categorical_features()
+        cat_features_from_backend = [self.fit_columns_renaming.get(c, c) for c in cat_features_from_backend]
         client_cat_features, search_keys_for_metrics = self._get_and_validate_client_cat_features(
             estimator, validated_X, search_keys
         )
@@ -1390,13 +1386,13 @@ class FeaturesEnricher(TransformerMixin):
                 cat_features_from_backend = [
                     c
                     for c in cat_features_from_backend
-                    if self.fit_columns_renaming.get(c, c) not in self.id_columns_encoder.feature_names_in_
+                    if c not in self.id_columns_encoder.feature_names_in_
                 ]
             if client_cat_features:
                 client_cat_features = [
                     c
                     for c in client_cat_features
-                    if self.fit_columns_renaming.get(c, c) not in self.id_columns_encoder.feature_names_in_
+                    if c not in self.id_columns_encoder.feature_names_in_
                 ]
 
         prepared_data = self._prepare_data_for_metrics(
@@ -1431,20 +1427,6 @@ class FeaturesEnricher(TransformerMixin):
             eval_set_dates,
         ) = prepared_data
 
-        # rename cat_features
-        if client_cat_features:
-            for new_c, old_c in columns_renaming.items():
-                if old_c in client_cat_features:
-                    client_cat_features.remove(old_c)
-                    client_cat_features.append(new_c)
-            for cat_feature in client_cat_features:
-                if cat_feature not in fitting_X.columns:
-                    self.logger.error(
-                        f"Client cat_feature `{cat_feature}` not found in" f" x columns: {fitting_X.columns.to_list()}"
-                    )
-        else:
-            client_cat_features = []
-
         model_task_type = self.model_task_type or define_task(y_sorted, has_date, self.logger, silent=True)
         cat_features = list(set(client_cat_features + cat_features_from_backend))
 
@@ -1496,14 +1478,6 @@ class FeaturesEnricher(TransformerMixin):
         # Find latest eval set or earliest if all eval sets are before train set
         date_column = self._get_date_column(search_keys)
 
-        if (
-            date_column is None
-            or not eval_set
-            or not eval_set_dates
-            or (self.cv is not None and self.cv.is_time_series())
-        ):
-            return []
-
         # Get minimum date from main dataset X
         main_min_date = X[date_column].min()
 
@@ -1539,15 +1513,29 @@ class FeaturesEnricher(TransformerMixin):
 
         checking_eval_set_df[date_column] = eval_set_dates[selected_eval_set_idx]
 
+        psi_values_sparse = calculate_sparsity_psi(
+            checking_eval_set_df, cat_features, date_column, self.logger, model_task_type
+        )
+
+        unstable_by_sparsity = [feature for feature, psi in psi_values_sparse.items() if psi > stability_threshold]
+        if unstable_by_sparsity:
+            self.logger.info(f"Unstable by sparsity features: {sorted(unstable_by_sparsity)}")
+
         psi_values = calculate_features_psi(
             checking_eval_set_df, cat_features, date_column, self.logger, model_task_type
         )
 
+        unstable_by_value = [feature for feature, psi in psi_values.items() if psi > stability_threshold]
+        if unstable_by_value:
+            self.logger.info(f"Unstable by value features: {sorted(unstable_by_value)}")
+
         self.psi_values = {
             feature: psi_value for feature, psi_value in psi_values.items() if psi_value <= stability_threshold
         }
 
-        return [feature for feature, psi in psi_values.items() if psi > stability_threshold]
+        total_unstable_features = sorted(set(unstable_by_sparsity + unstable_by_value))
+
+        return total_unstable_features
 
     def _update_shap_values(self, trace_id: str, df: pd.DataFrame, new_shaps: Dict[str, float], silent: bool = False):
         renaming = self.fit_columns_renaming or {}
@@ -1757,7 +1745,7 @@ class FeaturesEnricher(TransformerMixin):
     def _get_and_validate_client_cat_features(
         self, estimator: Optional[Any], X: pd.DataFrame, search_keys: Dict[str, SearchKey]
     ) -> Tuple[Optional[List[str]], List[str]]:
-        cat_features = None
+        cat_features = []
         search_keys_for_metrics = []
         if (
             estimator is not None
@@ -1926,7 +1914,7 @@ class FeaturesEnricher(TransformerMixin):
             fitting_X, y_sorted, search_keys, self.model_task_type, sort_all_columns=True, logger=self.logger
         )
         fitting_X = fitting_X[fitting_x_columns]
-        fitting_X, _ = self._encode_id_columns(fitting_X, self.fit_columns_renaming)
+        fitting_X, _ = self._encode_id_columns(fitting_X)
         self.logger.info(f"Final sorted list of fitting X columns: {fitting_x_columns}")
         fitting_enriched_x_columns = fitting_enriched_X.columns.to_list()
         fitting_enriched_x_columns = sort_columns(
@@ -1938,7 +1926,7 @@ class FeaturesEnricher(TransformerMixin):
             logger=self.logger,
         )
         fitting_enriched_X = fitting_enriched_X[fitting_enriched_x_columns]
-        fitting_enriched_X, _ = self._encode_id_columns(fitting_enriched_X, self.fit_columns_renaming)
+        fitting_enriched_X, _ = self._encode_id_columns(fitting_enriched_X)
         self.logger.info(f"Final sorted list of fitting enriched X columns: {fitting_enriched_x_columns}")
         date_column = self._get_date_column(search_keys)
         eval_set_dates = {}
@@ -1970,8 +1958,8 @@ class FeaturesEnricher(TransformerMixin):
                         .astype(np.float64)
                     )
 
-            fitting_eval_X, unknown_dict = self._encode_id_columns(fitting_eval_X, self.fit_columns_renaming)
-            fitting_enriched_eval_X, _ = self._encode_id_columns(fitting_enriched_eval_X, self.fit_columns_renaming)
+            fitting_eval_X, unknown_dict = self._encode_id_columns(fitting_eval_X)
+            fitting_enriched_eval_X, _ = self._encode_id_columns(fitting_enriched_eval_X)
 
             if len(unknown_dict) > 0:
                 print(self.bundle.get("unknown_id_column_value_in_eval_set").format(unknown_dict))
@@ -2299,13 +2287,7 @@ class FeaturesEnricher(TransformerMixin):
             enriched_df, x_columns, enriched_X.columns.tolist(), len(eval_set) if has_eval_set else 0
         )
 
-        # Add hash-suffixes because output of transform has original names
-        reversed_renaming = {v: k for k, v in columns_renaming.items()}
-        X_sampled.rename(columns=reversed_renaming, inplace=True)
-        enriched_X.rename(columns=reversed_renaming, inplace=True)
-        for _, (eval_X_sampled, enriched_eval_X, _) in eval_set_sampled_dict.items():
-            eval_X_sampled.rename(columns=reversed_renaming, inplace=True)
-            enriched_eval_X.rename(columns=reversed_renaming, inplace=True)
+        search_keys = {columns_renaming.get(k, k): v for k, v in search_keys.items()}
 
         # Cache and return results
         datasets_hash = hash_input(validated_X, validated_y, eval_set)
@@ -3205,7 +3187,7 @@ if response.status_code == 200:
             is_numeric_dtype(df[self.TARGET_NAME])
             and self.model_task_type in [ModelTaskType.BINARY, ModelTaskType.MULTICLASS]
             and has_date
-            and not self.cv.is_time_series()
+            and (self.cv is None or not self.cv.is_time_series())
         ):
             self._validate_PSI(df.sort_values(by=maybe_date_column))
 
@@ -3238,8 +3220,7 @@ if response.status_code == 200:
         self.fit_generated_features = [f for f in self.fit_generated_features if f not in self.fit_dropped_features]
 
         # Group columns should have normalized names
-        self.cv = None
-        self.__adjust_cv(df)
+        self.__adjust_cv(df, force=True)
         if self.id_columns is not None and self.cv is not None and self.cv.is_time_series():
             id_columns = self.__get_renamed_id_columns()
             if id_columns:
@@ -3544,19 +3525,21 @@ if response.status_code == 200:
         reverse_renaming = {v: k for k, v in renaming.items()}
         return None if self.id_columns is None else [reverse_renaming.get(c) or c for c in self.id_columns]
 
-    def __adjust_cv(self, df: pd.DataFrame):
+    def __adjust_cv(self, df: pd.DataFrame, force: bool = False):
+        if self.cv is not None and not force:
+            return
+
         date_column = SearchKey.find_key(self.fit_search_keys, [SearchKey.DATE, SearchKey.DATETIME])
         # Check Multivariate time series
         if (
-            self.cv is None
-            and date_column
+            date_column
             and self.model_task_type == ModelTaskType.REGRESSION
             and len({SearchKey.PHONE, SearchKey.EMAIL, SearchKey.HEM}.intersection(self.fit_search_keys.keys())) == 0
             and is_blocked_time_series(df, date_column, list(self.fit_search_keys.keys()) + [TARGET])
         ):
             msg = self.bundle.get("multivariate_timeseries_detected")
             self.__override_cv(CVType.blocked_time_series, msg, print_warning=False)
-        elif self.cv is None and self.model_task_type != ModelTaskType.REGRESSION:
+        elif self.model_task_type != ModelTaskType.REGRESSION:
             msg = self.bundle.get("group_k_fold_in_classification")
             self.__override_cv(CVType.group_k_fold, msg, print_warning=self.cv is not None)
             group_columns = self._get_group_columns(df, self.fit_search_keys)
@@ -3594,39 +3577,32 @@ if response.status_code == 200:
     def _encode_id_columns(
         self,
         X: pd.DataFrame,
-        columns_renaming: Optional[Dict[str, str]] = None,
     ) -> Tuple[pd.DataFrame, Dict[str, List[Any]]]:
-        columns_renaming = columns_renaming or {}
         unknown_dict = {}
 
         if self.id_columns and self.id_columns_encoder is not None:
-            inverse_columns_renaming = {v: k for k, v in columns_renaming.items()}
-            renamed_id_columns = [
-                inverse_columns_renaming.get(col, col) for col in self.id_columns_encoder.feature_names_in_
-            ]
-            self.logger.info(f"Convert id columns to int: {renamed_id_columns}")
-            encoded = self.id_columns_encoder.transform(X[renamed_id_columns].rename(columns=columns_renaming))
-            for i, c in enumerate(renamed_id_columns):
-                unknown_values = X[encoded[:, i] == -1][c].unique().tolist()
-                if len(unknown_values) > 0:
-                    unknown_dict[c] = unknown_values
-            X[renamed_id_columns] = encoded
-            X = X.loc[(X[renamed_id_columns] != -1).all(axis=1)]
-
-            if len(unknown_dict) > 0:
-                self.logger.warning(f"Unknown values in id columns: {unknown_dict}")
+            encoding_id_columns = [c for c in self.id_columns if c in X.columns]
+            if len(encoding_id_columns) > 0:
+                self.logger.info(f"Convert id columns to int: {encoding_id_columns}")
+                encoded = self.id_columns_encoder.transform(X[encoding_id_columns])
+                for i, c in enumerate(encoding_id_columns):
+                    unknown_values = X[encoded[:, i] == -1][c].unique().tolist()
+                    if len(unknown_values) > 0:
+                        unknown_dict[c] = unknown_values
+                X[encoding_id_columns] = encoded
+                X = X.loc[(X[encoding_id_columns] != -1).all(axis=1)]
+
+                if len(unknown_dict) > 0:
+                    self.logger.warning(f"Unknown values in id columns: {unknown_dict}")
 
         return X, unknown_dict
 
-    def _decode_id_columns(self, X: pd.DataFrame, columns_renaming: Dict[str, str]):
-        columns_renaming = columns_renaming or {}
+    def _decode_id_columns(self, X: pd.DataFrame):
         if self.id_columns and self.id_columns_encoder is not None:
-            inverse_columns_renaming = {v: k for k, v in columns_renaming.items()}
-            renamed_id_columns = [
-                inverse_columns_renaming.get(col, col) for col in self.id_columns_encoder.feature_names_in_
-            ]
-            decoded = self.id_columns_encoder.inverse_transform(X[renamed_id_columns].rename(columns=columns_renaming))
-            X[renamed_id_columns] = decoded
+            decoding_id_columns = [c for c in self.id_columns if c in X.columns]
+            if len(decoding_id_columns) > 0:
+                decoded = self.id_columns_encoder.inverse_transform(X[self.id_columns])
+                X[self.id_columns] = decoded
 
         return X
 
@@ -4172,7 +4148,7 @@ if response.status_code == 200:
         columns_to_sort = [date_column] if date_column is not None else []
 
         do_sorting = True
-        if self.id_columns and self.cv.is_time_series():
+        if self.id_columns and self.cv is not None and self.cv.is_time_series():
             # Check duplicates by date and id_columns
             reversed_columns_renaming = {v: k for k, v in columns_renaming.items()}
             renamed_id_columns = [reversed_columns_renaming.get(c, c) for c in self.id_columns]

upgini/utils/psi.py

@@ -42,6 +42,32 @@ DEFAULT_FEATURES_PARAMS = StabilityParams(
 )
 
 
+def calculate_sparsity_psi(
+    df: pd.DataFrame,
+    cat_features: list[str],
+    date_column: str,
+    logger: logging.Logger,
+    model_task_type: ModelTaskType,
+    psi_features_params: StabilityParams = DEFAULT_FEATURES_PARAMS,
+    psi_target_params: StabilityParams = DEFAULT_TARGET_PARAMS,
+) -> Dict[str, float]:
+    sparse_features = df.columns[df.isna().sum() > 0].to_list()
+    if len(sparse_features) > 0:
+        logger.info(f"Calculating sparsity stability for {len(sparse_features)} sparse features")
+        sparse_df = df[sparse_features].notna()
+        sparse_df[date_column] = df[date_column]
+        return calculate_features_psi(
+            sparse_df,
+            cat_features,
+            date_column,
+            logger,
+            model_task_type,
+            psi_target_params,
+            psi_features_params,
+        )
+    return {}
+
+
 def calculate_features_psi(
     df: pd.DataFrame,
     cat_features: list[str],

{upgini-1.2.113a3.dist-info → upgini-1.2.113a5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: upgini
-Version: 1.2.113a3
+Version: 1.2.113a5
 Summary: Intelligent data search & enrichment for Machine Learning
 Project-URL: Bug Reports, https://github.com/upgini/upgini/issues
 Project-URL: Homepage, https://upgini.com/

{upgini-1.2.113a3.dist-info → upgini-1.2.113a5.dist-info}/RECORD

@@ -1,9 +1,9 @@
-upgini/__about__.py,sha256=8dzTgZycBqXQF8H6x1n5BSvRmuPEQB4RCActXRN2QB8,26
+upgini/__about__.py,sha256=QdA0r4M8wEBY37BMjK9uA_83s1sWkyXy2XJhfn7vl3A,26
 upgini/__init__.py,sha256=LXSfTNU0HnlOkE69VCxkgIKDhWP-JFo_eBQ71OxTr5Y,261
 upgini/ads.py,sha256=nvuRxRx5MHDMgPr9SiU-fsqRdFaBv8p4_v1oqiysKpc,2714
 upgini/dataset.py,sha256=xFi0a-A3uvtxVwFM6JOyitkEPd1I2slIBj5SWfys3hQ,32724
 upgini/errors.py,sha256=2b_Wbo0OYhLUbrZqdLIx5jBnAsiD1Mcenh-VjR4HCTw,950
-upgini/features_enricher.py,sha256=oEVZVbN6qkeeTQkQfBCwnWw7hBEW9ZO2I9_r_m3bFUw,235890
+upgini/features_enricher.py,sha256=wifdmDP-3e3y51KYhCHPYuN6vU8mj2m3SYo-kMWcNz0,234523
 upgini/http.py,sha256=zeAZvT6IAzOs9jQ3WG8mJBANLajgvv2LZePFzKz004w,45482
 upgini/metadata.py,sha256=sx4X9fPkyCgXB6FPk9Rq_S1Kx8ibkbaWA-qNDVCuSmg,12811
 upgini/metrics.py,sha256=O19UqmgZ6SA136eCYV5lVU3J26ecgZlGXnxGblMvZJc,45869
@@ -64,7 +64,7 @@ upgini/utils/mstats.py,sha256=u3gQVUtDRbyrOQK6V1UJ2Rx1QbkSNYGjXa6m3Z_dPVs,6286
 upgini/utils/phone_utils.py,sha256=IrbztLuOJBiePqqxllfABWfYlfAjYevPhXKipl95wUI,10432
 upgini/utils/postal_code_utils.py,sha256=5M0sUqH2DAr33kARWCTXR-ACyzWbjDq_-0mmEml6ZcU,1716
 upgini/utils/progress_bar.py,sha256=N-Sfdah2Hg8lXP_fV9EfUTXz_PyRt4lo9fAHoUDOoLc,1550
-upgini/utils/psi.py,sha256=gYeZ3FOwGriVUnuO3BbdicSgXGQqPU14f7yleyO55f0,10108
+upgini/utils/psi.py,sha256=pLtECcCeco_WRqMjFnQvhUB4vHArjHtD5HzJFP9ICMc,10972
 upgini/utils/sample_utils.py,sha256=lZJ4yf9Jiq9Em2Ny9m3RIiF7WSxBPrc4E3xxn_8sQk8,15417
 upgini/utils/sklearn_ext.py,sha256=jLJWAKkqQinV15Z4y1ZnsN3c-fKFwXTsprs00COnyVU,49315
 upgini/utils/sort.py,sha256=8uuHs2nfSMVnz8GgvbOmgMB1PgEIZP1uhmeRFxcwnYw,7039
@@ -72,7 +72,7 @@ upgini/utils/target_utils.py,sha256=i3Xt5l9ybB2_nF_ma5cfPuL3OeFTs2dY2xDI0p4Azpg,
 upgini/utils/track_info.py,sha256=G5Lu1xxakg2_TQjKZk4b5SvrHsATTXNVV3NbvWtT8k8,5663
 upgini/utils/ts_utils.py,sha256=26vhC0pN7vLXK6R09EEkMK3Lwb9IVPH7LRdqFIQ3kPs,1383
 upgini/utils/warning_counter.py,sha256=-GRY8EUggEBKODPSuXAkHn9KnEQwAORC0mmz_tim-PM,254
-upgini-1.2.113a3.dist-info/METADATA,sha256=cyLDSMCN6fY7o3yXNpqMgPYG4TgY9CQNWA27O-ACBF8,49531
-upgini-1.2.113a3.dist-info/WHEEL,sha256=1yFddiXMmvYK7QYTqtRNtX66WJ0Mz8PYEiEUoOUUxRY,87
-upgini-1.2.113a3.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
-upgini-1.2.113a3.dist-info/RECORD,,
+upgini-1.2.113a5.dist-info/METADATA,sha256=VOeoK4hhJyhb0OJWG2cgsN-hES6xe3QIRyZMovxP8ek,49531
+upgini-1.2.113a5.dist-info/WHEEL,sha256=1yFddiXMmvYK7QYTqtRNtX66WJ0Mz8PYEiEUoOUUxRY,87
+upgini-1.2.113a5.dist-info/licenses/LICENSE,sha256=5RRzgvdJUu3BUDfv4bzVU6FqKgwHlIay63pPCSmSgzw,1514
+upgini-1.2.113a5.dist-info/RECORD,,